This assignment focus on the implementation of RCTs in social science in general and in economics, in particular. Specifically, we chose a list of articles which are based on RCT and appear in the Dataverse. Each parit would work on one of the articles, summarize it and replicate some parts of it using R software. In order to do so please follow the instructions below.
Stage I: Article Choice
Stage II: Article Summary (up to 8 pages, double space, font size 12)
- Summarize the article, following all of its sections (introduction, literature review etc.)
Pay special attention to the RCT design. Write in details what is the dependent variable, the independent variable (=the treatment effect), the controls, how the authors chose the sample ect. (It would be useful to follow the ‘general layout for an experiment design diagram’ we used in the first home assignment).
- Write in detail how the authors convince the readers that the design they chose allows them to infer about the causal effect of the independent variable on the dependent variable.
Stage III: Article Examination (up to 2 pages, double space, font size 12)
- Did the authors convince you that the RCT has internal validity? What do you think about the external validity of the results?
- If you could conduct your own RCT to answer the research questions the article aimed to investigate, what would you improve in the RTC design?
Stage VI: Replication
Download the dataset of your article (in right hand side of the title). Here you have data of the RCT + Stata files which allow you to replicate the regressions, tables etc.).
- Replicate the table of the Summary statistics of the main variables used in the RCT for the treatment and for the control group.
For example :
Ashraf, Nava, James Berry, and Jesse M. Shapiro. 2010. “Can Higher Prices Stimulate Product Use? Evidence from a Field Experiment in Zambia.” American Economic Review, 100 (5): 2383-2413. doi: 10.1257/aer.100.5.2383
- Replicate the main regressions of the article in R (choose 3 regressions if there more than 3 regressions). In the appendix attach the original regressions from the article
Good Luck!